This work deals with interaction control of a linear motor in a HiL test rig that should be able to interact with objects with high contact stiffness. The mathematical model of the linear motor includes nonlinear friction. In the course of a theoretical survey, common control algorithms are assessed for their practicability in a mechanical HiL environment. Impedance- and admittance control concepts are presented with their advantages and disadvantages. The implemented interaction control algorithm consists of two control loops. An adaptive admittance control strategy is developed, which utilizes online estimation of the environment dynamics to preserve the desired dynamic behaviour independently of environment stiffness and sampling time. A self-perturbing recursive least squares estimator is implemented for environment identification. For the inner position control loop, a reference model following controller with friction feedforward is chosen. It is shown that a state observer delivers unsatisfactory results due to static friction. Instead a sensor fusion approach is implemented to obtain a virtual velocity measurement with improved accuracy. The proposed approach shows good performance as well as promising stability behaviour, while posing minimal constraints on the choice of virtual environment dynamics.